Digital transformation or digitalization is widely discussed today. Digitalization offers an abundance of products based on new technologies and technological platforms that possess the capability to reshape business processes, organizational structure and ways of working. It holds the power to reshape complete business models or even overall market situations.
One main element of digitalization is cloud computing, the usage of IT resources (CPU time, storage, etc.) only up to the amount that is necessary at a certain point in time. Enterprise architecture management (EAM) needs to include the resource used in the cloud (be it private, public or a community cloud) to prevent the IT landscape from spontaneous growth. Measures on how to govern the usage of cloud services should (or need to be) implemented. Today, we’d like to show you how.
Get Architecture Information from AWS
Cloud services like Microsoft Azure, OpenStack or Amazon Web Services (AWS) offer APIs by which many different cloud resources can be created and managed.
We had a closer look at AWS. Its API allows control of instances (virtual servers), virtual machine images, volumes, hosts or network infrastructures, etc. For our first implementation, we concentrated on the instances, to be exact, on the AWS Elastic Cloud Compute (EC2) instances. These are the resource most likely used by our customers – i.e. moving applications from an on-premise server to instances in the cloud.
We created a command line tool that runs regularly to request the instance information from the cloud provider and download it in a standardized format.
Integrate with Enterprise Architecture Management
Based on that an automated import job is started to draw the information in Alfabet. Alfabet is a powerful EAM tool that we use frequently in customer EAM projects. Alfabet provides an object type called “device” that represents a server where an application can be deployed on. We used this object type to describe the AWS instances in the realm of Alfabet:
The import job also connects the imported instances to the existing IT landscape:
- Instances are connected to the location where they are running (i.e. “eu-central-1” for the AWS data center in Frankfurt, Germany).
- Users are assigned to the instances to manage them.
- A workflow is run to ask responsible users what applications are running on the instances. This is necessary to link instances and applications, and thus make clear how instances are used for.
The steps above provide transparency about the cloud IT landscape. They are the preliminary work for the “actual” use cases:
- One could review all running applications whether they can be deployed on a cloud instance or not. This would enhance IT standardization as cloud services are built on standardized hardware and software platforms.
- One could import and collect cost information per instance and use this for reporting and budgeting purposes.
Presentation at EAMKON 2017
We will talk about the integration of the Alfabet EAM tool with AWS at the EAMKON 2017 conference in Stuttgart, 30th May 2017. Looking forward to seeing you there!